242 REPORTS OF INVESTIGATIONS AND PROJECTS. 



The experiments with guinea-pigs have been concerned, in part, with the 

 effects of selection upon variations in size occurring within a single race of 

 guinea-pigs, that described in Publication No. 49, Carnegie Institution of 

 Washington. This experiment has now been in progress for two years, and 

 is in immediate charge of a research student, Mr. C. C. Little. By selecting 

 at birth, from the same litters of young, the largest and the smallest individu- 

 als, respectively, it has been possible to establish two subraces distinctly 

 different in average size, though still intergrading. It remains to differentiate 

 these subraces, if possible, until they no longer intergrade, and then to test 

 the stability of the distinction established between the two races. The poly- 

 dactylous character of this family of guinea-pigs (see Publication No. 49), 

 which was originally built up by selection, has fully maintained itself through 

 the several generations which have intervened since selection for this char- 

 acter was suspended. 



The hybridization experiments between the female guinea-pig and the 

 male Cavia aperea have this year made good progress. In these unquestion- 

 able "true species crosses" the same laws of heredity govern as in crosses 

 between different varieties of guinea-pig. The same sharp Mendelian domi- 

 nance and segregation of color-characters occur among the hybrids as among 

 guinea-pigs, while the polydactylous character shows the same imperfection 

 of dominance and segregation, and skeleton dimensions and proportions are 

 as in guinea-pigs fully blending. 



The half-blood hybrids resulting from a first cross are fertile in the female 

 sex, but without observed exception sterile in the male sex ; the same is true 

 of the quarter-blood aperea hybrids resulting from crosses of the female half- 

 bloods with the male guinea-pig, so far as yet tested. One-eighth blood and 

 one-sixteenth blood apereas are now being produced in considerable numbers, 

 and it will be interesting to see whether fertile males will occur among these. 

 With each successive cross since the first the wild disposition of the aperea 

 occurs in diminished intensity, indicating that this does not have its basis in 

 a simple Mendelian character. Selection experiments for modification of 

 Mendelian color-characters by processes other than mutation are being con- 

 tinued with apparent success. Evidence for the existence of Mendelian units 

 is not thereby weakened, but the fact is shown that these units are modifiable. 

 The experiments with rats have been directed chiefly toward the same 

 theoretical question, the modifiability of Mendelian characters. It has been 

 shown that by selection the extent of the pigmented areas on piebald rats may 

 be gradually and permanently altered. Two races of "hooded" rats have now 

 been established which no longer intergrade, while in one of these the extent 

 of the pigmentation has been so increased that this series now intergrades 

 with the "Irish," a variety originally quite distinct from the hooded, and 

 characterized by pigmentation much more extensive. In this investigation I 

 have been fortunate to have the assistance of Dr. John C. Phillips, research 



